1278
Biochemistry 2006, 45, 1278-1284
A Thaumatin-like Protein from Larvae of the Beetle Dendroides canadensis Enhances the Activity of Antifreeze Proteins† Lei Wang and John G. Duman* Department of Biological Sciences, UniVersity of Notre Dame, Notre Dame, Indiana 46556 ReceiVed August 23, 2005; ReVised Manuscript ReceiVed December 2, 2005
ABSTRACT: The levels of thermal hysteresis (antifreeze activity) produced by purified antifreeze proteins (DAFPs) from the larvae of the beetle Dendroides canadensis at endogenous concentrations are lower than what are present in the hemolymph of overwintering larvae. Thermal hysteresis activity of DAFPs is dependent not only on AFP concentration but also on the presence of enhancers that may be either proteins (including other hemolymph DAFPs) or low-molecular mass enhancers such as glycerol. The purpose of this study was to identify endogenous protein enhancers using yeast two-hybrid, coimmunoprecipitation, and finally the enhancement of antifreeze activity. Here we show that a thaumatinlike protein from D. canadensis, until recently known only from plants, significantly enhances the thermal hysteresis of DAFP-1 and -2. Glycerol can further this enhancement, presumably by promoting the interaction of the DAFPs and thaumatin-like protein.
Antifreeze proteins (AFPs)1 lower the nonequilibrium freezing point of water while not affecting the melting point. This results in a difference between the freezing and melting points that is termed thermal hysteresis (1). According to the adsorption-inhibition mechanism of action (2), the AFPs depress the freezing point by adsorbing onto the surface of ice crystals at preferred growth sites (2-4), by means of hydrogen bonding (5, 6), hydrophobic, and/or van der Waals interactions (7-9). This adsorption permits ice crystal growth to occur only between the AFPs in highly curved (high surface free energy) fronts rather than the preferred lowcurvature (low surface free energy) fronts, and therefore, the temperature must be lowered below the equilibrium melting point before ice crystal growth can take place. Overwintering larvae of the beetle Dendroides canadensis produce a family of antifreeze proteins (DAFPs) that function to inhibit inoculative freezing across the cuticle (10) and to promote supercooling by inhibiting ice nucleators (11-13). The group I DAFPs (1, 2, 4, and 6) are present in the hemolymph, while groups II and III DAFPs occur in the gut fluid (14). There are both low-molecular mass (15) and highmolecular mass antifreeze protein enhancers (16) present in the hemolymph, and these significantly increase the thermal hysteresis activity (THA). The protein enhancers interact with the DAFPs, forming complexes that, because they are larger than the DAFP alone, block a larger surface area of the † This work was supported by National Science Foundation Grant IBN02-12907 and by the Indiana 21st Century Fund for Research and Technology. * To whom correspondence should be addressed. Telephone: (574) 631-9499. Fax: (574) 631-7413. E-mail:
[email protected]. 1 Abbreviations: AFP, antifreeze protein; DAFP, D. canadensis antifreeze protein; THA, thermal hysteresis activity; PCR, polymerase chain reaction; cDNA, complementary deoxyribonucleic acid; SDS, sodium dodecyl sulfate; PAGE, polyacrylamide gel electrophoresis; HA, hemagglutinin; IPTG, isopropyl β-D-thiogalactoside; OD, optical density; TLP-Dcan, D. canadensis thaumatin-like protein.
potential seed ice crystal than does the DAFP alone. Therefore, THA is increased. Because the protein enhancers are thought to bind to the DAFPs, a yeast two-hybrid screen was used to identify D. canadensis proteins that interact with DAFPs and might therefore be potential enhancers. This screen previously demonstrated that certain group I DAFPs (1, 2, 4, and 6) found in the hemolymph interact with one another, and these were then shown to enhance the thermal hysteresis activities of one another (17). The study reported here demonstrates that a thaumatin-like protein (until recently known only from plants) was also identified by the yeast two-hybrid screen as interacting with DAFP-1 and -2. The recombinant thaumatin-like protein was subsequently shown to enhance DAFP-1 and -2 activities. EXPERIMENTAL PROCEDURES Rationale. This project used a three-step approach to identify D. canadensis proteins that interact with DAFPs and enhance DAFP activity. Initially, the yeast two-hybrid system was used to screen a D. canadensis cDNA library for “prey” proteins which interact with DAFP “bait”. This was followed by co-immunoprecipitation to verify that proteins identified by the yeast two-hybrid screen interact with certain DAFPs. Finally, proteins identified by these two procedures as interacting with DAFPs were expressed in an Escherichia coli expression system, purified, and then tested for their ability to enhance certain DAFPs by comparing the thermal hysteresis of aqueous solutions of DAFPs with and without the potential enhancer. Yeast Two Hybrid Screen. The Matchmaker Two-Hybrid System 3 (CLONETECH Laboratories, Inc.) used in this study is a GAL4-based two-hybrid system that provides a transcriptional assay for detecting protein-protein interaction in vivo in yeast cells. (See Matchmaker GAL4 Two-Hybrid System 3 & Libraries User Manual.) A bait gene (encoding
10.1021/bi051680r CCC: $33.50 © 2006 American Chemical Society Published on Web 01/07/2006
Thaumatin-like Protein Enhances Antifreeze Proteins certain DAFPs) was expressed as a fusion protein with the GAL4 DNA-binding domain, while a cDNA (encoding a possible prey protein) is expressed as a fusion with the GAL4 activation domain. When bait and library fusion prey proteins interact, the DNA-binding domain and the activation domains are brought into proximity, thereby activating transcription of four reporter genes. Construction of Bait Plasmids. As described previously (17), the cDNA encoding the mature proteins of D. canadensis DAFP-1 was obtained by PCR from the pBluescript plasmid containing the appropriate sequence using the following primers: 5′ primers, 5′ GCTGAATTCCAATGTACTGGTGGTTCCGAT 3′; and 3′ primer, 5′ GCAGGATCCTGGACATCCCGTGGAATCGGT 3′. The sequence was designed to include an EcoRI restriction site in the 5′ primer and a BamHI restriction site in the 3′ primer. Amplified DNA was cleaved with the restriction site enzymes and cloned into pGBKT7 that had been previously linearized with the same enzymes. The resulting plasmids, which contain the DAFP-1 coding sequences in frame with the DNA-binding domain of Gal4, were designated pGBKT7DAFP-1. These constructs were used as baits for screening a Dendroides whole-body cDNA library. Western Blot Analysis. To verify that DAFP-1 is expressed well in yeast strain AH109, pGBKT7-DAFP-1 was transformed into strain AH109, and the protein extracts were prepared from yeast by the urea-SDS method (Clontech yeast protocols handbook). The Western blot was carried out with a polyclonal antibody (made using DAFP-1 and -2 as antigens) to DAFPs using purified DAFP-1 as a positive control and plain yeast strain AH109 protein extracts as a negative control. Construction of the Prey cDNA Library [As Described PreViously (17)]. The D. canadensis larvae were collected in late October. Total RNA was extracted from 3 g of D. canadensis whole body, and part of it was used to extract mRNA and then synthesize the cDNA (Stratagene cDNA synthesis kit) with a designed EcoRI restriction site in the 5′ primer and an XhoI restriction site in the 3′ primer. These cDNAs were then cloned into the pGADT7 vector that was lineanized with the same enzymes. The primary titer of the constructed cDNA library was 1.5 × 107 pfu/mL, and the titer of the amplified library was 2.4 × 1010 pfu/mL. The average size of inserts was 1.2 kb. The resulting plasmids, which contained the cDNAs of D. canadensis whole body with the DNA activation domain of Gal4, were designated the pGADT7 library. Yeast Two-Hybrid Screen [As Described PreViously (17)]. The AH109 yeast strain was simultaneously transformed using large-scale transformation with pGBKT7-DAFP-1 and the pGADT7 library using the lithium acetate method (Clontech BD Matchmaker Two-Hybrid System 3). The transformants were plated on synthetic medium lacking adenine, histidine, leucine, and tryptophan, and with X-RGal. Colonies were picked at times from 4 to 8 days and replated on the second-generation plates for 6-10 days with synthetic medium lacking histidine, leucine, and tryptophan, and with X-RGal. The blue colonies that were well-isolated were picked up and replated on synthetic medium lacking adenine, histidine, leucine, and tryptophan to verify that they maintained the correct phenotype. The yeast plasmids were purified using the lyticase method (Clontech yeast protocols
Biochemistry, Vol. 45, No. 4, 2006 1279 handbook), transformed into E. coli using electroporation, and purified using a Miniprep kit (Qiagen). The purified pGAD plasmids were retransformed into strain AH109 carrying pGBKT7-DAFP-1 to verify positive colonies. The positives were sequenced by automated DNA sequencing (ABI 3700). In Vitro Transcription/Translation and Immunoprecipitation. As described previously (17), the pGBKT7-DAFP-1 and pGAD-TLP-Dcan1 plasmids from the yeast two-hybrid system were transcribed and translated using Promega’s TNT T7 Coupled Reticulocyte Lysate System in vitro to prepare [35S]Met-labeled bait and prey proteins. The interaction between DAFP-1 and TLP-Dcan1 was confirmed and visualized by in vitro co-immunoprecipitation (Co-IP) using the BD Matchmaker Co-IP kit (Clontech) which is compatible with the BD Matchmaker Two-Hybrid System 3. The co-immunoprecipitates were separated by SDS-PAGE and analyzed by autoradiography. 5′ Rapid Amplificaiton of cDNA Ends (RACE). 5′ RACE, followed by automated sequencing, was used to determine the full-length sequence of cDNAs encoding proteins that were shown by yeast two-hybrid and co-immunoprecipitation to interact with DAFPs. The RNA Oligo ligated to full-length mRNA was made from the same total RNA with which the pGADT7 library was made. Then the RACE-ready cDNA was made by reverse-transcribing the mRNA; this was used as a template for 5′ RACE (Invitrogen). Two approaches were used to amplify and identify the 5′ end of TLP-Dcan1. In the first approach, the primary amplification was performed on 0.3 µL of cDNA template using GeneRacer 5′ primers (5′ CGACTGGAGCACGAGGACACTGA 3′) and TLP-Dcan1 gene-specific 3′ primers (5′ GCGAAAGCTTGGCACGTGAAGGTACTCTTGT 3′). For secondary amplification, 1 µL of reactants was taken from the first amplification product and amplified using GeneRacer 5′ Nested primers (5′ GGACACTGACATGGACTGAAGGAGTA 3′) and TLP-Dcan1 gene-specific 3′ primers (5′ GCGAAAGCTTGGCACGTGAAGGTACTCTTGT 3′), and PCR products were identified by automated sequencing (ABI 3700). Expression of Recombinant DAFPs and TLP-Dcan1. Recombinant DAFP-1, -2, -4, and -6 were expressed in E. coli BL21(DES) pLysS competent cells as described previously (17). The DNA encoding the mature TLP-Dcan1 was amplified by PCR using the 5′ primer (5′ TCGGATCCTTCGTGCTCCTAGTC 3′) and the 3′ primer (5′ TAGTGCGGCCGCGCACGTGAAGGTACTCTTGTGGTC 3′). Sequences were designed to include a BamHI restriction site in the 5′ primer and an NotI restriction site in the 3′ primer. Amplified DNA was digested and cloned into pET-20b (Novagen), previously linearized with the same enzymes. The resulting plasmids were sequenced (ABI 3700). They were then transformed into E. coli BL21(DE3) pLysS competent cells. The cells were grown in LB broth containing carbenicillin, ampicillin, and chloramphenicol. IPTG was added to the culture to a final concentration of 0.4 mM for pet-20b(+) when the OD600 reached 0.6. Cells were grown at 37 °C for 3 h in the induction medium and collected by centrifugation at 6000 rpm for 20 min. The extracted crude protein was purified using the His-bind purification kit (Novagen). The purified protein was dialyzed, freeze-dried, and redissolved in sterilized distilled water.
1280 Biochemistry, Vol. 45, No. 4, 2006 RT-PCR of TLP-Dcan1. RT-PCR was used to determine when the transcript encoding TLP-Dcan1 was produced by the larvae. Total RNA was isolated from both fat body cells and gut epithelium cells of winter and summer D. canadensis. First-strand cDNA was synthesized using the First Strand cDNA synthesis kit (Novagen). The cDNA products were amplified using the TLP-Dcan1-specific 5′ primer (5′ CCA TCT CAG CAA CGG TGG TCT CAT CCT CGA CC 3′) and the 3′ primer (5′ GGT AGT CGA CGG GCC AGT CCG AAG ACT TGC AG 3′). Thermal Hysteresis ActiVity. The capillary technique was used to measure thermal hysteresis activity (1). The sample (∼4 µL) was placed in a sealed capillary tube; a small (∼0.25 mm) seed crystal was spray-frozen in the sample, and the capillary was placed into a refrigerated alcohol bath which was equipped with a viewing chamber through which the crystal could be observed with a microscope. The bath temperature was controlled to (0.01 °C. The temperature was slowly increased (0.02 °C/5 min) until the crystal disappeared. This temperature was taken as the melting point of the sample. Another crystal was spray-frozen in the sample, the capillary again placed in the chamber, and the bath temperature lowered (0.1 °C/2.5 min) until the crystal grew. The temperature at which the seed crystal began to grow was taken as the hysteretic freezing point. Aqueous solutions, which do not contain thermal hysteresis antifreeze proteins, exhibit identical freezing and melting points. However, in the presence of these proteins, the freezing point is often depressed several degrees below the melting point. This difference between the melting and freezing points is termed thermal hysteresis. Generally, at least two freezing/ melting point determinations were made for each sample, and the two measurements were usually within e0.4 °C of one another. If the difference was greater, additional measurements were taken. RESULTS To search for cDNA clones encoding proteins that interact with DAFP-1, the coding sequence for D. canadensis hemolymph DAFP-1 was fused with the Gal4 DNA-binding domain and used to screen a D. canadensis whole-body library expressed from the pGADT7 vector. Nineteen clones were isolated that displayed an Ade+His+LacZ+ phenotype, representing 15 different cDNAs as judged by cDNA sequencing. However, only three turned out to be positives on the basis of co-immunoprecipitation results. As reported previously (17), two of these were other DAFPs (DAFP-2 and DAFP-4). The interaction of DAFP-1 with these proteins was specific, as no interaction was observed between the identified proteins and the Gal4 DNA-binding domain (BD) alone. The interaction was confirmed by co-immunoprecipitation (Figure 1). One positive yeast clone contained the partial cDNA sequence of a thaumatin-like protein. Using a RACE approach, the complete gene was sequenced. The open reading frame encoded a 253-amino acid protein (Figure 2) with a predicted molecular mass of 28 kDa. The N-terminal region of the protein presents a putative signal peptide of 15 amino acids. The 28 kDa mature protein consists of 253 amino acids. Its sequence is very similar to those of thaumatin-like
Wang and Duman
FIGURE 1: Interaction between DAFP-1 and TLP-Dcan1 was confirmed by co-immunoprecipitation (Co-IP). The pGBKT7DAFP-1 and pGAD-TLP-Dcan1 plasmids from the yeast twohybrid system were transcribed and translated as [35S]Met-labeled Myc-tagged DAFP-1 and [35S]Met-labeled HA-tagged TLP-Dcan1. Either Myc monoclonal or HA-Tag polyclonal antibodies were used to precipitate the mixture of [35S]Met-labeled Myc-taged DAFP-1 and [35S]Met-labeled HA-tagged TLP-Dcan1. The co-immunoprecipitates were separated by SDS-PAGE and analyzed by autoradiography: lane 1, [35S]Met-labeled Myc-tagged DAFP-1 and [35S]Met-labeled HA-tagged TLP-Dcan1 mixture precipitated by HA-Tag polyclonal antibodies; lane 2, [35S]Met-labeled Myc-tagged DAFP-1 precipitated by Myc monoclonal antibodies.
proteins in insects and plants (Figure 3). It is 65% identical and 76% positive with TLP of Schistocercus gregaria (TLPSgre1, GenBank accession number AY512591) and 47% identical and 61% positive with TLP of Arabidopsis thaliana (TLP-Atal, GenBank accession number NP177641). Reverse transcription using TLP-Dcan1-specific primers was done to confirm the expression of the TLP-Dcan1 gene. The expected amplification products were present in both the fat body cells and gut epithelial cells, in both winter and summer (Figure 4). However, in summer the fat-body amplification product was larger than expected (Figure 4, lane 3), and by sequencing all the amplification products, we identified another TLP in D. canadensis, TLP-Dcan2. Although the TLP-Dcan2 transcript is larger than that encoding TLP-Dcan1 (Figure 4), the former contains a stop codon at residue 525 and consequently TLP-Dcan2 has a lower molecular mass than TLP-Dcan1. The open reading frame encodes a 177-amino acid protein (Figure 5) with a predicted molecular mass of 19 kDa, including a putative signal peptide of 15 amino acids. To determine if the interaction between DAFP-1 and TLPDcan1 can enhance the thermal hysteresis activity of DAFP1, mature DAFP-1 and TLP-Dcan1 were expressed in the bacterial expression system and purified and the THA of DAFP-1 alone was compared with that of DAFP-1 to which various amounts of purified TLP-Dcan1 were added. Figure 6 shows that the thermal hysteresis activities of the 0.5 and 1.0 mg/mL DAFP-1 solution were enhanced by addition of TLP-Dcan1, even at very low TLP-Dcan1 concentrations. For example, at a concentration of 0.5 mg/mL TLP-Dcan1, the THA of the 0.5 mg/mL DAFP solution was increased from 2.7 to ∼7 °C. At TLP-Dcan1 concentrations above 0.5 mg/mL, the THA of the 0.5 mg/mL DAFP solution did not increase further. In similar experiments, the addition of TLPDcan1 to DAFP-2 resulted in enhancement of THA; however, no enhancement was seen with addition of TLPDcan1 to the other two hemolymph DAFPs (4 and 6) (data not shown). TLP-Dcan1 alone does not exhibit THA. As shown in Figure 7, the enhancement of DAFP-1 by TLP-Dcan1 is further enhanced by glycerol. Glycerol, which
Thaumatin-like Protein Enhances Antifreeze Proteins
Biochemistry, Vol. 45, No. 4, 2006 1281
FIGURE 2: Nucleotide and deduced amino acid sequence of the 28 kDa TLP-Dcan1. The asterisk indicates the predicted signal peptide cleavage site (between residues 15 and 16).
FIGURE 3: Complete amino acid sequence of the thaumatin-like protein (TLP) of D. canadensis (TLP-Dcan1, GenBank accession number DQ023319), aligned with representative TLP of S. gregaria (TLP-Sgre1, GenBank accession number AY512591), C. elegans (TLP-Cele, GenBank accession number NP502362), and A. thaliana (TLP-Atal, GenBank accession number NP177641). Identical residues are shaded.
is present in the hemolymph of D. canadensis in winter at concentrations of 0.5-1.0 M, was especially effective at low TLP-Dcan1 concentrations. For example, at 0.25 mg/mL TLP-Dcan1, the THA of 0.5 mg/mL DAFP-1 increased from 3.5 to 8.7 °C. However, there was much less enhancement at the higher TLP-Dcan1 concentrations. It was previously shown that DAFP-1 and -2 enhance one another and that glycerol provides further enhancement (17). Therefore, 1 mg/mL TLP-Dcan1 was added to a solution containing DAFP-1 and -2 in equal concentrations with 0.5 M glycerol. As demonstrated in Figure 8, the addition of
TLP-Dcan1 greatly increased THA, at both high and low DAFP concentrations. However, comparison of Figures 7 and 8 indicates that the addition of DAFP-2 (Figure 8) to the combination of DAFP-1, TLP-Dcan1, and glycerol (Figure 7) did not increase the maximal THA. All four group I DAFPs (1, 2, 4, and 6) are normally present in the hemolymph. Consequently, TLP-Dcan1 was added to a mixture of equal concentrations of DAFP-1, -2, -4, and -6 with glycerol (Figure 9). As seen previously (17), these four DAFPs enhance one another at higher, but not lower, DAFP concentrations (>1 mg/mL). In contrast, the
1282 Biochemistry, Vol. 45, No. 4, 2006
FIGURE 4: RT-PCR showing the presence of the TLP-Dcan1 transcript in fat body cells and gut epithelium cells in both winter and summer. First-strand cDNA synthesized from total RNA of winter and summer fat body cells and gut epithelium cells were amplified using TLP-Dcan1-specific primers: lane 1, 100 bp marker (Promega); lane 2, winter whole body; lane 3, summer fat body; lane 4, winter fat body; lane 5, summer gut; and lane 6, winter gut. Note the slower migration of the transcript in lane 3.
addition of TLP-Dcan1 resulted in a significant enhancement at the lower, but not the higher, DAFP concentrations. In fact, at the highest DAFP concentration that was tested (3 mg/mL), the addition of TLP-Dcan1 and glycerol to the mixture of the four DAFPs actually resulted in less THA. DISCUSSION Thaumatin, a sweet-tasting 207-residue protein, was first isolated from the fruits of the West African rain forest shrub Thaumatococcus daniellii; however, proteins with similar amino acid sequences have been identified in all plant species that have been examined (18-20). Pathogenesis-related (PR) proteins are produced by plants in response to environmental stress signals such as pathogen invasion, drought, and wounding. Members of family 5 of the PR proteins (PR-5 proteins) consist of thaumatin-like proteins (TLPs) and osmotin-like proteins. All PR-5 proteins have considerable sequence similarities, and they all have antifungal and/or antibacterial functions, whether they are classified as thaumatin-like or osmotin-like proteins. (Osmotin is a salt-stressinduced protein originally described from tobacco.) The structure of thaumatin has been determined (21, 22). The TLPs are generally resistant to proteases and pH- or heatinduced denaturation. This is likely due to the presence of 16 conserved cysteines that form eight disulfide bridges. Their action against pathogenic microorganisms has been related in some cases to their endo-β-1,3-glucanase activity (23, 24) and to R-amylase inhibiting properties (25). Also, PR-5 proteins have been reported to have membranepermeabilizing activity (26).
Wang and Duman On the basis of genome information, thaumatin-like proteins were predicted to exist in the nematode Caenorhabditis elegans (27). Recently, for the first time in animal species, genes encoding members of the thaumatin-like protein family were reported in the desert locust S. gregaria and in the related Locusta migratoria (28). The threedimensional model of the S. gregaria thaumatin-like protein suggests a glucanase function, but this has not been confirmed experimentally. It has been shown that endogenous protein and lowmolecular mass enhancers (glycerol) are present and play important roles in increasing the thermal hysteresis activity in the hemolymph of overwintering D. canadensis larvae (16). Since protein enhancers of DAFPs, including polyclonal antibodies specific for DAFPs, are known to bind to the DAFPs (29, 30), the yeast two-hybrid system was used to identify endogenous D. canadensis proteins which interact with certain DAFPs and which therefore might be endogenous DAFP enhancers. Among the proteins identified by the yeast two-hybrid screen were other DAFPs (17) as well as a thaumatin-like protein (TLP-Dcan1). Binding between DAFP-1 and TLP-Dcan1 was confirmed by co-immunoprecipitation. The interaction was further confirmed by adding TLP-Dcan1 to DAFP-1, resulting in enhanced THA (Figure 6). For example, with addition of 0.5 mg/mL TLPDcan1, the THA of a 0.5 mg/mL DAFP-1 solution was increased from 2.7 to ∼7 °C. Our interpretation is that the DAFP-1-TLP-Dcan1 complex can still bind to ice, but the complexes are able to block a larger surface area of the ice and/or are more difficult for the ice to overgrow; consequently, the THA is increased. However, at TLP-Dcan1 concentrations above 0.5 mg/mL, the enhancement did not increase further (Figure 6). This could be a result of saturated binding of TLP-Dcan1 to DAFP-1. TLP-Dcan1 had a similar enhancing effect on DAFP-2. Glycerol was previously shown to enhance the activity of DAFPs (15), presumably by stimulating enhanced DAFP interactions with one another (17). Glycerol also appears to promote the interaction between DAFP-1 and TLP-Dcan1, which consequently further improved the ability of TLPDcan1 to enhance the activity of DAFP-1 (Figure 7). Figure 10 demonstrates the interactions of the four hemolymph (group I) DAFPs. DAFP-1 interacts with DAFP-2 and -4 but not DAFP-6, even though DAFP-4 and -6 are very similar (sequence differences at just four positions).
FIGURE 5: Complete amino acid sequence of the thaumatin-like protein (TLP) of D. canadensis (TLP-Dcan1, GenBank accession number DQ023319), aligned with TLP-Dcan2 (GenBank accession number DQ112346). Different amino acids are shown in TLP-Dcan2 using lowercase letters.
Thaumatin-like Protein Enhances Antifreeze Proteins
FIGURE 6: Effect of TLP-Dcan1 on THA of a 0.5 or 1.0 mg/mL DAFP-1 solution. The thermal hysteresis activities of the 0.5 and 1.0 mg/mL DAFP-1 solutions were enhanced by addition of TLPDcan1: (0) TLP-Dcan1 and 0.5 mg/mL DAFP-1 and (4) TLPDcan1 and 1.0 mg/mL DAFP-1. Values are means of two or three measurements of the same sample. The standard deviation ranged from 0.00 to 0.21. Computer curve fitting (Excel) was used to connect points in the graph to guide the eye.
FIGURE 7: Enhancement of DAFP-1 by TLP-Dcan1 is further promoted by glycerol, especially at the lower TLP-Dcan1 concentrations. TLP-Dcan1 from 0 to 1 mg/mL was added to the solution with 0.5 mg/mL DAFP-1 and 0.5 M glycerol or with 1.0 mg/mL DAFP-1 and 0.5 M glycerol: (0) 0.5 mg/mL DAFP-1, TLP-Dcan1, and 0.5 M glycerol, (9) TLP-Dcan1 and 0.5 mg/mL DAFP-1, (2) 1.0 mg/mL DAFP-1 and TLP-Dcan1, and (4) 1.0 mg/mL DAFP1, TLP-Dcan1, and 0.5 M glycerol. Values are means of two or three measurements of the same sample. The standard deviation ranged from 0.00 to 0.28. Computer curve fitting (Excel) was used to connect points in the graph to guide the eye.
DAFP-4 interacts with DAFP-1 and -6 but not DAFP-2, even though DAFP-1 and -2 are different at only two positions. Although DAFP-1 and -2 enhance one another, and glycerol provides further enhancement (17), comparison of Figures 7 and 8 indicates that the combination of DAFP-1 and DAFP-2 with TLP-Dcan1 did not increase the THA over that of TLP-Dcan1 with DAFP-1 alone, either with or without glycerol. This is interesting since TLP-Dcan1 is capable of enhancing either DAFP-1 or -2 alone. Also, TLPDcan1 has no effect on the THA of either DAFP-4 or DAFP-6 (data not shown). However, as shown in Figure 9, the combination of all four group I hemolymph DAFPs, with glycerol, was considerably enhanced by TLP-Dcan1 at lower DAFP concentrations (up to 1 mg/mL). The composition(s), size(s), and shape(s) of these complexes are not known at this time. Also, the composition, size, and stability of the complexes could vary with protein concentration (as suggested by Figure 9), protein ratios, or other factors. Consequently, this information must be gathered before an informed theory of the mechanism of increased freezing point
Biochemistry, Vol. 45, No. 4, 2006 1283
FIGURE 8: Effect of TLP-Dcan1 on the thermal hysteresis activity of the combination of DAFP-1 and -2: (2) DAFP-1, (O) DAFP-2, (4) DAFP-1 and DAFP-2, (9) DAFP-1, DAFP-2, and 0.5 M glycerol, and (0) DAFP-1, DAFP-2, 0.5 M glycerol, and 1.0 mg/ mL TLP-Dcan1. Values are means of two or three measurements of the same sample. The standard deviation ranged from 0.00 to 0.28. Computer curve fitting (Excel) was used to connect points in the graph to guide the eye.
FIGURE 9: Effect of TLP-Dcan1 on the thermal hystresis activity of a combination of DAFP-1, -2, -4, and -6: (]) DAFP-1, (O) DAFP-2, (0) DAFP-4, (×) DAFP-6, (2) DAFP-1, DAFP-2, DAFP4, DAFP-6, and 0.5 M glycerol, and (9) DAFP-1, DAFP-2, DAFP4, DAFP-6, 1.0 mg/mL TLP-Dcan1, and 0.5 M glycerol. Values are means of two or three measurements of the same sample. The standard deviation ranged from 0.00 to 0.28. Computer curve fitting (Excel) was used to connect points in the graph to guide the eye.
FIGURE 10: Interactions of the four hemolymph (group I) DAFPs (summarized from ref 17 and this study).
depression of these complexes can be put forward. However, these data are likely to have important implications for the understanding of the mechanism of action of AFPs. Group I DAFPs (1, 2, 4, and 6) are produced by fat body cells and secreted into the hemolymph (14). TLP-Dcan1 is also produced by the fat body, and since TLP-Dcan1 has a signal peptide, it is probably also secreted into the hemolymph. RT-PCR demonstrated that gut epithelial cells produce TLPDcan1 in the winter and TLP-Dcan2 in the summer (Figure 4). Because of the presence of a signal peptide, both are apparently secreted into the midgut fluid. While DAFP-1 is not present in the midgut fluid, group II and III DAFPs are produced in the gut epithelia and secreted into the gut fluid
1284 Biochemistry, Vol. 45, No. 4, 2006 (14). At this time, we do not know if TLP-Dcan1 enhances any of the gut DAFPs; however, when DAFP-8 (group III) was used as bait in a yeast two-hybrid screen, neither of the TLP-Dcan proteins was identified as a positive (L. Wang and J. G. Duman, unpublished observations). Despite this, it is possible that TLP-Dcan1, which is produced by gut epithelia in winter (Figure 4), enhances one or more of the gut DAFPs. However, at this time, the function of TLPDcan1 in the gut fluid remains unknown. Also, since TLPDcan1 is produced in the fat body in summer as well as in winter (Figure 4), it is possible that it has another function (i.e., antimicrobial activity) in addition to enhancing DAFPs. Multiple functions of plant PR proteins are not uncommon. In fact, a number of plant antifreeze proteins are PR proteins (including some TLPs), having a sequence similar to those of PR proteins and providing protection against psychrophilic microbial pathogens (31). Since the plant AFPs produce only minimal thermal hysteresis (a few tenths of a degree, at most) and appear to function to inhibit the recrystallization of ice in intracellular spaces of the freeze tolerant plants that produce them, it is perhaps preferable to refer to them as ice-binding proteins, rather than AFPs. In any event, although as mentioned earlier, TLP-Dcan1 lacks AFP activity, it is interesting that TLP-Dcan 1 has evolved to enhance DAFPs. REFERENCES 1. DeVries, A. L. (1986) Antifreeze glycopeptides and peptides: interactions with ice and water, Methods Enzymol. 127, 293303. 2. Raymond, J. A., and DeVries, A. L. (1977) Adsorption inhibition as a mechanism of freezing resistance in polar fishes, Proc. Natl. Acad. Sci. U.S.A. 86, 881-885. 3. Raymond, J. A., Wilson, P. W., and DeVries, A. L. (1989) Inhibition of growth on nonbasal planes in ice by fish antifreeze, Proc. Natl. Acad. Sci. U.S.A. 86, 881-885. 4. Knight, C. A., Cheng, C. C., and DeVries, A. L. (1991) Adsorption of R-helical antifreeze peptides on specific ice crystal surface planes, Biophys. J. 59, 409-418. 5. DeVries, A. L., and Cheng, C. H. C. (1992) The role of antifreeze glycopeptides and peptides in the survival of cold-water fishes, in Water and Life: ComparatiVe Analysis of Water Relationships at the Organismic, Cellular and Molecular LeVel (Somero, G. N., and Osmond, C. B., Eds.) pp 301-315, Springer-Verlag, Berlin. 6. Sincheri, F., and Yang, D. S. C. (1995) Ice-binding structure and mechanism of an antifreeze protein from winter flounder, Nature 375, 427-431. 7. Cheng, A., and Merz, K. M. (1997) Ice-binding mechanism of winter flounder antifreeze proteins, Biophys. J. 73, 2851-2873. 8. Chao, H., Houston, M. E., Hodges, R. S., Kay, C. M., and Sykes, B. D. (1997) A diminished role for hydrogen bonds in antifreeze binding to ice, Biochemistry 36, 14652-14660. 9. Haymet, A. D., Ward, L. G., Harding, M. M., and Knight, C. A. (1998) Valine substituted winter flounder ‘antifreeze’: Preservation of ice growth hysteresis, FEBS Lett. 430, 301-306. 10. Olsen, T. M., Sass, S., Li, N., and Duman, J. G. (1998) Factors contributing to seasonal increases in inoculative freezing resistance in overwintering fire-colored beetle larvae Dendroides canadensis, J. Exp. Biol. 201, 1585-1594. 11. Olsen, T. M., and Duman, J. G. (1997) Maintenance of the supercooled state in overwintering Pyrochroid beetle larvae
Wang and Duman Dendroides canadensis: Role of hemolymph ice nucleators and antifreeze proteins, J. Comp. Physiol., B 167, 105-113. 12. Olsen, T. M., and Duman, J. G. (1997) Maintenance of the supercooled state in the gut of overwintering Pyrochroid beetle larvae Dendroides canadensis: Role of gut ice nucleators and antifreeze proteins, J. Comp. Physiol., B 167, 114-122. 13. Duman, J. G. (2002) The inhibition of ice nucleators by insect antifreeze proteins is enhanced by glycerol and citrate, J. Comp. Physiol., B 172, 163-168. 14. Duman, J. G., Verleye, D., and Li, N. (2002) Site-specific forms of antifreeze protein in the beetle Dendroides canadensis, J. Comp. Physiol., B 172, 547-552. 15. Li, N., Andorfer, C. A., and Duman, J. G. (1998) Enhancement of insect antifreeze protein activity by solutes of low molecular mass, J. Exp. Biol. 201, 2243-2251. 16. Duman, J. G., and Serianni, A. S. (2002) The role of endogenous antifreeze protein enhancers in the hemolymph thermal hysteresis activity of the beetle Dendroides canadensis, J. Insect Physiol. 48, 103-111. 17. Wang, L., and Duman, J. G. (2005) Antifreeze proteins of beetle Dendroides canadensis enhance one another’s activities, Biochemistry 44, 10305-10312. 18. Cornelissen, B. J., Hooft van Huijsduijnen, R. A., and Bol, J. F. (1986) A tobacco mosaic virus-induced tobacco protein is homologous to the sweet-tasting protein thaumatin, Nature 321, 531-532. 19. Van Loon, L. C., Gerritsen, Y. A. M., and Ritter, C. E. (1987) Identification, purification and characterization of pathogenesisrelated proteins from virus-infected Samsun NN tobacco leaves, Plant Mol. Biol. 9, 593-609. 20. Breiteneder, H., and Ebner, C. (2000) Molecular and biochemical classification of plant-derived food allergens, J. Allergy Clin. Immunol. 106, 27-36. 21. Ogata, C. M., Gordon, P. F., de Vos, A. M., and Kim, S. H. (1992) Crystal structure of a sweet tasting protein thaumatin I, at 1.65 Å resolution, J. Mol. Biol. 228, 893-908. 22. Kaneko, R., and Kitabatake, N. (2001) Structure-sweetness relationship in thaumatin: Importance of lysine residues, Chem. Senses 26, 167-177. 23. Grenier, J., Potvin, C., Trudel, J., and Asselin, A. (1999) Some thaumatin-like proteins hydrolyse polymeric β-1,3-glucans, Plant J. 19, 473-480. 24. Menu-Bouaouiche, L., Vriet, C., Peumans, W. J., Barre, A., Van Damme, E. J., and Rouge, P. (2003) A molecular basis for the endo-β-1,3-glucanase activity of the thaumatin-like proteins from edible fruits, Biochimie 85, 123-131. 25. Franco, O. L., Rigden, D. J., Melo, F. R., and Grossi-De-Sa, M. F. (2002) Plant R-amylase inhibitors and their interaction with insect R-amylases, Eur. J. Biochem. 269, 397-412. 26. Koiwa, H., Kato, H., Nakatsu, T., Oda, J., Yamada, Y., and Sato, F. (1999) Crystal structure of tobacco PR-5d protein at 1.8 Å resolution reveals a conserved acidic cleft structure in antifungal thaumatin-like proteins, J. Mol. Biol. 286, 1137-1145. 27. Kitajima, S., and Sato, F. (1999) Plant pathogeniesis-related proteins: Molecular mechanisms of gene expression and protein function, J. Biochem. 125, 1-8. 28. Brandazza, A., Angeli, S., Tegoni, M., Cambillau, C., and Pelosi, P. (2004) Plant stress proteins of the thaumatin-like family discovered in animals, FEBS Lett. 572, 3-7. 29. Wu, D. W., and Duman, J. G. (1991) Activation of antifreeze proteins from larvae of beetle Dendroides canadensis, J. Comp. Physiol., B 161, 279-283. 30. Wu, D. W., Duman, J. G., and Xu, L. (1991) Enhancement of antifreeze protein activity by antibodies, Biochim. Biophys. Acta 1076, 416-420. 31. Griffith, M., and Yaish, M. W. F. (2004) Antifreeze proteins in overwintering plants: A tale of two activites, Trends Plant Sci. 9, 399-405. BI051680R
